Determination of sulphur trioxide in the presence of sulphur
Transcript of Determination of sulphur trioxide in the presence of sulphur
S5S4
DETERMINATION OF SULPHUR TRIOXIDE IN THEPRESENCE OF SULPHUR DIOXIDE, TOGETHER WITHSOME ANALYSES OF COMMERCIAL LIQUID SULPHURDIOXIDE
By Jo R. Eckman
ABSTRACT
A method for determining sulphur trioxide in sulphur dioxide has been de-
veloped for use in a study of the heat of formation of sulphur dioxide. It mayfind application also in control work in the manufacture of sulphur dioxide,
sulphites, and sulphuric acid, and in the analysis of oxygen-free products of
combustion of fuels. None of the methods given in the literature appeared
suitable for the purpose. Therefore a method has been devised for the complete
removal of sulphur trioxide by precipitation as barium sulphate in a slightly
acid solution of barium sulphite, and filtration and washing of the precipitate
with oxygen-free distilled water in an atmosphere of oxygen-free nitrogen. Themethod has been applied to the analysis of eight samples of commercial liquid
sulphur dioxide.
CONTENTSPage
I. Introduction 277
II. Sulphur trioxide in gaseous sulphur dioxide 279
1. Apparatus 279
2. Procedure 280
3. Efliciency of method 282
4. Possible simplification of method and apparatus 283
TIL Sulphur trioxide in Hquid sulphur dioxide : 284
1. Apparatus 2842. Procedure 284
3. Results 285IV. Summary 285
I. INTRODUCTION
The analytical method described in this paper was developed in
connection with a study of the heat of formation of sulphur dioxide.
It is published at this time on account of its possible application in
control work in the manufacture of sulphur dioxide, sulphites, andsulphuric acid, and in the analysis of oxygen-free products of com-bustion of fuels. The literature contains a number of different
methods which have been used for this determination. Berthelot ^ in
his work on the heat of formation of sulphur dioxide burned a definite
amount of sulphur and absorbed the gas in the amount of iodine solu-
tion that would be required if all of the sulphur were burned to sulphur
1 M. Berthelot, "Sur la chaleur de formation des oxydes du soufre," Ann. chim. [5], 22, p. 422; 1881.
'—27 277
278 Scientific Papers of the Bureau of Standards i voi n
dioxide. Any unused iodine was regarded as a measure of the sulphur
trioxide present in the gas. Thomsen ^ condensed the sulphur tri-
oxide on the walls of the platinum combustion chamber, washed it
out with water and titrated. Ljungh,^ using the principle of the
Reich-Lunge method/ determiaed dioxide plus trioxide by absorption
in standard sodium hydroxide, and the dioxide by absorption in stand-
ard iodine solution. Pellet ^ absorbed the trioxide in concentrated
hydrochloric acid and determined it gravimetrically by precipitation
with barium chloride. Hawiey ^ used a sheet of filter paper moistened
with water to collect the trioxide and then titrated the resulting
sulphuric acid. Richter ^ collected the trioxide in condenser tubes
and determined it in the usual gravimetric way. Dunn ^ investigated
the Reich-Lunge method * of estimating sulphur trioxide by difference
and found it inaccurate. In his own method he absorbed trioxide
plus dioxide in barium chloride solution containing 10 to 15 per cent
of hydrochloric acid, removed the dioxide by boiling and deter-
mined the trioxide in the usual gravimetric way. Saillard and Breg-
mann ^ used an electrical conductivity method. Dieckmann ^°
absorbed the gases in sodium hydroxide and determined the relative
amounts of trioxide and dioxide by acid titration using double indi-
cators. Sander ^^ used standard sodium hydroxide as an absorbent
for trioxide plus dioxide, added mercuric chloride to liberate a
quantity of hydrochloric acid equivalent to the dioxide, and then
titrated with standard sodium hydroxide. Gille ^^ collected the
trioxide in a plug of cotton. Schmidt ^^ used as absorbents iodine
solutions and varied the concentration for the purpose of getting
selective absorption of the trioxide and dioxide.
None of these methods appeared suitable for the present purpose
and a method comprising the following essential features was ac-
cordingly devised. The sulphur trioxide is removed from the gas
* Julius Thomsen, " Thermochemische Untersuchungen," Johann A. Barth, Leipzig, Z, p. 245; 1882.
'H. Ljungh, "ZurBestimmungdesSOj-Verlustes in den Abgasen einer Kontaktschwefelsaurefabrik,"
Chem.-Zeitg., S3, p. 143; 1909.
* Q. Lunge, "Technical Methods of Chemical Analysis," English Translation of 2d. German ed., Gurney& Jackson, London, 1, pp. 299-303; 1908.
» H. Pellet, " Sur le dosage de I'acide sulfureux et de I'acide sulfurique dans les gaz du four & soufre,"
Bull, assoc. chim. sucr. dist., 27, p. 468; 1909.
6 H. G. Hawiey, "Determining sulphur trioxide in flue gas," Eng. Min. J., 94, p. 987; 1913.
7 Erich Richter, "Uber die Bestimmungen von SOs in Schwefelofengasen," Z. angew. Chem., US, p. 132;
1913. Also Wochbl. Papierfabr., 54, p. 1521; 1923.
8 E. M. Dunn, "Determination of gases in smelter flues," Bull. Am. Inst. Min. Eng., 80, p. 2051; 1913.
' E. Saillard et Dr. Bregmann, " D6termination de I'acide sulfureux et de I'acide sulfurique
des gaz des fours a soufre par conductibilitS electrique," Rev. g§n. chim., 20, p. 69; 1917.
w R. Dieckmann, "Uber eine vereinfaehte Methode zur Bestimmung von Schwefeldioxyd und Schwe-feltrioxyd in den Rostgasen," Papierfabr., 19, p. 285; 1921. Also Chem.-Zeitg., 45, p. 885; 1921.
" A. Sander, "tJber die titrimetrische Bestimmung freier schwefliger SSure." Chem.-Zeitg., 38, p. 1057;
1914. " Eine neue Methode zur XJntersuchung der Rostgase und der Sulfitlauge," Papier-Zeitg., 45, p. 1842;
1920. "Uber die Bestimmung von Schwefeldioxyd neben Schwefeltrioxyd in Rostgasen und Oleiun,"
Chem.-Zeitg., 46, p. 261; 1921.
" H. Gille, " Uber Absorption chemischer Nebel," Z. angew. Chem., 39, p. 401; 1926.
" E. Schmidt, "Uber Rostgasanalyse," Papierfabr., 23, p. 229; 1925. "Die Bestimmung von Schwe-feltrioxyd in Rostgasen," Z. angew. Chem., 39, p. 732; 1926.
Scientific Papers of the Bureau of Standards, Vol. 22
Fig. 1.
—
Apparatus used, with modified form of analytical tube
Eckman] Suljphur THoxide in Sulphur Dioxide 279
stream by means of a special filtering device; the sulphur trioxide is
precipitated by a nearly saturated and slightly acid solution of bariumsulphite which must be free from oxygen; and the barium sulphate
is filtered and washed in an atmosphere of nitrogen free from oxygen.
Various difficulties and sources of error involved in the earlier meth-ods are thus eliminated.
II. SULPHUR TRIOXIDE IN GASEOUS SULPHUR DIOXIDE
1. APPARATUS
A photograph of the actual apparatus used with the modified formof analytical tube is shown in Figure 1. A diagrammatic sketch of
the apparatus is shown in Figure 2. ^ is a small metal cylinder
containing sulphur dioxide. B and D are glass safety traps. 6'' is a
glass flow meter containing water. £' is a bottle used for the prepa-
ration of the barium-sulphite reagent. This bottle is fitted with a
paraffined four-hole rubber stopper through which pass the sulphur
dioxide inlet tube 2-F, inlet or outlet tube o-G, inlet tube 4~^i, and
the barium-sulphite outlet tube H-5-g, which is equipped with either
an alundum filtering tip at H, or a plug of cotton at /, to hold back
any loose particles of precipitate which might otherwise be carried
over with the barium-sulphite reagent.
J is the analytical tube into which clear barium-sulphite reagent
from bottle E is forced for the purpose of absorbing the sulphur
trioxide. In the lower part of tube J is sealed a sintered-glass
filter plate, N, the porosity of which is suitable for filtering barium
sulphate by means of suction without danger of loss. Below this
plate is a two-way stopcock, 5, the stem, i?, of which extends through
a rubber stopper, j, into suction flask 8. Tube J is fitted at the
top with a groimd-glass stopper, li, in which are sealed three tubes,
one closed by two-way stopcocks 6 and 7, the second by three-way
stopcock 12, and the third by two-way stopcock 10. The inlet tube
12-M, through which the gas sample is admitted, ends in either an
alundum filtering tip, or a sintered-glass plate, M, for the purpose
of collecting any sulphur trioxide present as mist or fog.^^ TJ is an
inlet with ground-glass stopper for the introduction of hot hydro-
chloric acid. TF is a mercury gauge used to regulate the pressure
of the nitrogen admitted to the apparatus at e. The gas to be
analyzed is admitted through inlet/.
F is a flask for the preparation of oxygen-free distilled water. It
is fitted with a ground-glass stopper, fc, containing three tubes 17~Zf
18-Zi, and I9-Z2. Z is a funnel tube with a ground-glass stopper
through which distilled water may be introduced into flask Y,
" See footnote 13, p. 278.
280 Scientific Papers oj the Bureau oj Standards t Vol. gg
Ground-glass joints h, d, and g permit the removal of bottle Efor cleaning and refilling. The connection to cylinder A is made bythreaded brass joint a.
<^
O
pi.
a,
1
6
2. PROCEDURE
Through inlet o about 500 ml of 0.02 normal hydrochloric acid,
containing 5 g of barium chloride (BaCl2.2H20), is introduced into
bottle E. With stopcocks 4 and 5 closed and stopcock 3 open sulphur
Ecitman] SulpTiur THoxidc in SulpTiur Dioxide 281
dioxide from cylinder J. is run into bottle jE" through stopcocks 1 and 2and inlet F until the solution is saturated. Stopcocks 1 and 2 are
then closed. Nitrogen ^^ is passed in through stopcocks IS and 4and inlet ft and out through stopcock 3 in order to remove all the
oxygen from that line. Stopcock 4 is then closed and stopcock 3left open.
Nitrogen is then passed through stopcocks 15, 14j and 12 into tube
J and thence through stopcocks 6 and 5, opening H and stopcock 3,
Stopcock 3 is then closed. In a similar manner oxygen is driven
out of line N-8-R. In the initial determination lines K-7-16-17-Zj14-17-Zj and 14-18-Zi are thus freed from oxygen. Oxygen is
removed from line 11-12 by suction at/, and from line 20-16 in the
initial determination by suction at p.
Oxygen-free distilled water is prepared in flask Y by boiling dis-
tilled water with nitrogen bubbling through it by way of line 14-17-
Z-19.
Stopcocks 8 and 6 are closed and stopcocks 7 and 10 are opened
and, by means of nitrogen pressure exerted at inlet Zi through
stopcock 18, about 20 ml of the oxygen-free distilled water is forced
through line Z-17-16-7-K into tube J. With stopcocks 7 and 8
closed and stopcocks 10, 6, and 5 open, and with nitrogen entering
bottle E through stopcock 4 and inlet Gi, about 80 ml of the clear
barium-sulphite reagent is forced through line H-g-K into tube J,
which is thus filled to height X. Stopcocks 6 and 5 are then closed.
The sample of gas to be analyzed is then admitted at/ and drawn or
forced through stopcocks 20, 11, and 12 and tip if into the barium-
sulphite reagent. The amount of gas used can be determined by an
appropriate measuring device at outlet g. Any sulphur trioxide
which may have condensed in line 20-1 1-12-M is washed into tube Jwith oxygen-free distilled water admitted through stopcock 20. After
the precipitate has settled the solution is filtered by means of suction
applied at T through filter plate N into flask S containing one or two
drops of sulphuric acid for the purpose of showing whether the filtrate
contains excess barium. A stream of nitrogen admitted through
line 15-14-12-Mis aUowed to follow the effluent solution at approxi-
mately atmospheric pressiire as indicated on gauge W. Oxidation
of any sulphurous acid or barium sulphite is thus prevented. Theprecipitate is now washed several times, or until free from barium,
with oxygen-free distiUed water forced, by nitrogen pressure at inlet
Zi, from flask Y through line Z-17-16-7-K into tube J. The wash-
" Nitrogen from a cylinder was freed from oxygen by passing "it through copper-ammonia-ammonium
chloride reagent (W. L. Badger, J. Ind. Eng. Chem., 12, 161; 1920). Ammonia thus introduced into the
nitrogen was then absorbed in sulphuric acid (1:1). Any carbon dioxide present was removed by soda-lime,
and water vapor by "Dehydrite."
282 Scientific Papers of the Bureau oj Standards [ voi. n
ings are sucked through filter plate iV and stopcock 8 into flask iS' and
discarded.
Tube J may now be disconnected from suction flask S and from its
ground-glass stopper Ih. With the aid of a so-called "policeman*'
and ordinary distilled water from a wash bottle most of the bariumsulphate is transferred from tube J to a weighed platinum dish.
Any film of precipitate still adhering to the inner wall of tube J is
dissolved in boiling 2-normal hydrochloric acid. Any precipitate
formed in the tube above tip M from sulphur trioxide mist is likewise
treated with the boiling 2-normal hydrochloric acid introduced through
inlet JJ. The combined hot acid washings, amoimting to about
200 ml, are then drawn by suction through filter plate iV into a clean
suction flask and transferred to the weighed platinxmi dish contain-
ing the main portion of the barium sulphate, and the whole evaporated
to dryness. The residue is treated with two or three drops of dilute
sulphuric acid, heated until free from fumes, ignited over a TirrUi
burner, cooled in a desiccator, and weighed as barium sulphate. Asmall blank, due mainly to nonvolatile residue from the hydrochloric
acid and determined as explained below, is substracted. (Table 2.)
3. EFFICIENCY OF METHOD
Tests on analytical tube J were made to determine its efiiciency
in removing sulphur trioxide. For this purpose a tube similar to
analytical tube J was inserted between outlet n and inlet/. This
intermediate tube contained the same kind and depth of barium-
sulphite reagent as tube J . Sulphur dioxide from cylinder A wasthen bubbled, at the rate of 40 ml a minute for 90 minutes, by wayof stopcock 1 through this reagent and then by way of stopcocks
$0^ 11 J and 12 and tip M through the reagent in tube J. An appre-
ciable quantity of barium sulphate was precipitated in the intermediate
tube but none in tube J, indicating that the pro-
n cedure is satisfactory for the complete removal of
^ ^^ small amounts of sulphur trioxide.
I
^^ A stni more conclusive test of the method wasp-^ ^ then made as follows: Tube J was filled to height^^ L with the barium-sulphite reagent. A slow stream
Fig. Z.—Means of q£ nitrogen was admitted to the apparatus at e and
plur trioxide ^^^ through stopcocks 15, H, and 12 and tip J/ into
the barium-sulphite reagent. A known amount of
sulphur trioxide was then introduced into the nitrogen stream byheating 1 ml of standard sulphuric acid in bulb V (fig. 3), which hadbeen sealed into line 12-M and filled through inlet TJ by means of a 1
ml pipette graduated in hundredths of a milliliter. The remainder
of the procedure was the same as explained above under II, 2.
The results obtained are given in Table 1.
Ecltman] Sulphur Trioxide in Sulphur Dioxide
Table 1
283
No.StandardH2SO4 used
SO3 equiva-lent
SO3 found
1
ml±O.OOS1.0001.000
g0. 0412
. 0412
g0. 0409
2 .0411
Average . . .0410
Blanks on this test were run in a similar manner, except that nosulphur trioxide was introduced into the nitrogen stream. Theresults of these blanks are given in Table 2.
Table 2
No.Residue
weighed asBaS04
SO3 equiva-lent
1.. _ V0003.0003
g0.0001
2 .0001
4. POSSIBLE SIMPLIFICATION OF METHOD ANDAPPARATUS
.-^
S-^~
Some attempts were made to simplify the methodso as to avoid the necessity of carrying out the
operations in an atmosphere of nitrogen. All such
attempts failed, however, because of the rapidity
with which the barium-sulphite reagent undergoes
oxidation on contact with the atmosphere. The use
of an oxygen-free atmosphere is therefore necessary
if reliable results are to be obtained.
The analytical tube J may, however, be modified
if desired so as to avoid the necessity of using the
sintered-glass filter plate M. Such a modified formis shown in Figures 1 and 4. Below stopcock 8 (fig. 4)
is a tubular extension, P, with a small inner-sealed
tube ending at tip 0. Extension P is joined by a
heavy rubber band or a ground-glass joint, ^, to a
funnel-shaped tube, Q. The funnel is fitted with a
platiniun filtering cone and an ashless filter paper.
On the lower end of funnel Q is a two-way stopcock,
9, connected to suction flask S in the same way as in
the apparatus shown in Figure 2. Precipitation of
sulphur trioxide is accomplished in tube J, butfiltration is carried out in funnel Q. This modified
analytical tube does not, however, reduce the time
required for a determination, since hot hydrochloric acid mustbe used to wash the analytical tube and must be evaporated just as
in a determination carried out with analytical tube / shown in Figure 2.
J-
Fig. 4.
—
Modi-
fied form of
analytical tube
284 Scientific Papers of the Bureau oj Standards i Vol. 22
III. SULPHUR TRIOXIDE IN LIQUID SULPHUR DIOXIDE
1. APPARATUS
The apparatus used for sampling and distilling the liquid sulphur
dioxide is shown in Figure 5. ^ is a Pyrex flask with a narrow cali-
brated neck and ground-glass stopper fitted with inlet and outlet
tubes, each containing a three-way stopcock; s is a threaded brass
connector and w a ground-glass joint.
2. PROCEDURE
Flask A is filled with nitrogen and placed in a Dewar flask. Thecylinder of sulphur dioxide is shaken and then inverted. Flask A is
then connected to the needle valve on the cylinder by means of
joint s and filled with liquid sulphur dioxide to a point just above
mark x. The flask is disconnected and, with stopcock 21 closed andstopcock 22 open through by-pass v, it is connected to analytical tube
J (fig. 2) at inlet/ by means of ground-
glass joint w. Inlet r is sealed to anextension from d with de Khotinsky
cement. As soon as enough of the
sample has been distilled through by-
pass V to bring the meniscus to mark x,
stopcock 22 is turned so as to allow the
gas to pass through Ime f-20-11-12-
M
into the barium-sulphite reagent. Thepressure in flask A, as noted on a mer-
cury gauge sealed in the line between
outlet u and stopcock 22 j and the rate
of distillation may be regulated by rais-
ing or lowering the Dewar flask. Whendistillation is complete the residual sul-
phur dioxide gas in flaskA is flushed out
into the reagent with nitrogen. In the
case of the wet sample, 5, the residue contained some white crystals,
known as Pictet^s crystals and supposed to be a compound of sul-
phur dioxide and water. The sulphur dioxide was readily driven
out of these crystals by very gentle heating. The water left in the
flask was heated in a nitrogen stream until free from sulphur dioxide.
Any precipitate formed in the analytical tube is treated as ex-
plained above in the analysis of sulphur dioxide gas. In some of the
determinations the modified form of analytical tube shown in Figures
1 and 4 was used and found very satisfactory.
Sulphuric acid in the residue is determined by precipitation with
barium chloride in the usual gravimetric way.
Fig. 5.
—
Apparatus for sampling
and distilling liquid sulphur
dioxide
EcTiman] SulpTiur THoxide in Sulphur Dioxide 285
3. RESULTS
The following eight samples of liquid sulphur dioxide from com-mercial sources were analyzed
:
1. Specially purified and dried by fractionation.
2. Conmiercial product; partially purified and dried by fractiona-
tion.
3. Factory product; not fractionated to remove water.
4. Commercial product; dried with sulphuric acid.
5. Commercial product; dried with sulphuric acid.
6. Commercial product; dried with sulphuric acid.
7. Dried twice with sulphuric acid; slightly contaminated with oil
from compressor during second drying.
8. Same as 7.
All samples were taken from 35-pound lots except sample 6 which
came from a 5-pound cylinder. About 200 g was used for each
determination and complete distillation required from 2% to 4 hours.
In no case was any precipitate found in the analytical tube, showing
the gas to be free from sulphur trioxide. Only in samples 3,7, and 8
was any sulphur trioxide found in the residue. Kesults for the residue
from sample 3 could not be checked and varied from a trace to 0.0032
per cent of sulphur trioxide, indicating nonuniformity of sample. In
each of samples 7 and 8, 0.0011 per cent of sulphur trioxide was found.
IV. SUMMARY
1. A nearly saturated and slightly acid solution of barium sulphite
free from oxygen is used as the reagent for precipitating sulphur
trioxide.
2. A means is provided for filtering and washing the precipitated
barium sulphate with oxygen-free distilled water in an atmosphere
of oxygen-free nitrogen.
3. Results are given for ei^ht different samples of commercial
liquid sulphur dioxide.
The author wishes to express his gratitude to E. W. Washburn,
chief chemist of this bureau, for his able guidance and valuable sug-
gestions, which are largely responsible for the completion of this work.
Acknowledgment is made to A. K. Scribner, of the Virginia Smelt-
ing Co., for furnishing the samples of liquid sulphur dioxide and
assisting in their analysis. The author is indebted to S. L. Earle, of
the chemistry division, for making the drawings, and to E. O. Sperling,
glassblower of the bureau, for the construction of apparatus, and for
suggestions iu design.
Washington, March 21, 1927.